Inadvertent Transfer of Murine VL30 Retrotransposons to CAR-T Cells

For more than a decade, genetically engineered autologous T-cells have been successfully employed as immunotherapy drugs for patients with incurable blood cancers. The active components in some of these game-changing medicines are autologous T-cells that express viral vector-delivered chimeric antigen receptors (CARs), which specifically target proteins that are preferentially expressed on cancer cells. Some of these therapeutic CAR expressing T-cells (CAR-Ts) are engineered via transduction with γ-retroviral vectors (γ-RVVs) produced in a stable producer cell line that was derived from murine PG13 packaging cells (ATCC CRL-10686). Earlier studies reported on the copackaging of murine virus-like 30S RNA (VL30) genomes with γ-retroviral vectors generated in murine stable packaging cells. In an earlier study, VL30 mRNA was found to enhance the metastatic potential of human melanoma cells. These findings raise biosafety concerns regarding the possibility that therapeutic CAR-Ts have been inadvertently contaminated with potentially oncogenic VL30 retrotransposons. In this study, we demonstrated the presence of infectious VL30 particles in PG13 cell-conditioned media and observed the ability of these particles to deliver transcriptionally active VL30 genomes to human cells. Notably, VL30 genomes packaged by HIV-1-based vector particles transduced naïve human cells in culture. Furthermore, we detected the transfer and expression of VL30 genomes in clinical-grade CAR-T cells generated by transduction with PG13 cell-derived γ-retroviral vectors. Our findings raise biosafety concerns regarding the use of murine packaging cell lines in ongoing clinical applications.


Introduction
Short sequences of homology between the human genome and various VL30 genomes. The transcriptional activity of VL30 genomes in primary human T-cells and the ability of HIV-1 vector particles to mobilize VL30 genomes from human cells (Figure 4) increase the potential for recombination between VL30 genomes and either human or/and HIV-1 genomes. [28,[32][33][34] Prompted by this possibility, we searched for sequence homologies between various VL30 genomes and either the HIV-1 genome or the human genome. We could not detect sequence homology between VL30 genomes and the HIV-1 genome. Importantly, multiple short sequences of ~30 bp in VL30 genomes were determined to be homologous to sequences throughout the human genome. Most of the abovementioned VL30-homologous human sequences were composed of interspersed repeats, including LTRs, LINEs, and SINEs, or simple repeats (Table  7).

Discussion
Following successful clinical trials, in 2017 [35,36] the US Food and Drug Administration (FDA) approved Tisagenlecleucel (Kymriah® Novartis) and axicabtagene ciloleucel (Yescarta® Kite Pharma) as the first two gene therapy-based immunotherapy drugs for patients with hematologic malignancies. The active component of the novel anticancer drugs is made of autologous T-Cells that express viral vector-delivered chimeric antigen receptors (CARs). The CAR-Ts in both drugs target the B cell CD19 protein. Initially, Kymriah® and Yescarta® were indicated for young (under 25 years old) patients with acute lymphocytic leukemia (ALL) or adult patients with large B-cell lymphoma respectively, who relapsed or did not respond to two conventional anticancer treatments. [35,36] HIV-1-based vectors generated by transient four-plasmid transfection of human cells deliver the CAR expression cassette to Kymriah® CAR-Ts. Generation of Yescarta® CAR-Ts is mediated by transduction with the g-RVV PG13-CD19-H3. The vector is produced in a stable producer cell line, which was isolated by Kochenderfer et al. [5] as a single cell clone (clone H3) following transduction of the murine PG13 packaging cells [6] (ATCC CRL-10686). Prior to and since 2017, the PG13 packaging cell line and its derivatives have been employed to generate g-RVV-transduced T-cells for various anticancer protocols. [37][38][39][40][41][42][43][44][45][46]. Importantly, all tested g-RVVs employed in the abovementioned clinical trials were determined to be free of replicationcompetent retroviruses (RCRs). [38,40] In 2020, the FDA and the European Medicines Agency (EMA) approved a second medicine premised on g-RVV-transduced CAR-T cells, Tecratus, as an immunotherapy medicine. [47,48] Tecratus is indicated for the treatment of adult patients with relapsed/refractory mantle cell lymphoma (MCL). The g-RVVs employed to generate Tecratus and Yescarta® CAR-T cells are identical and are generated in the same PG13 producer cell line. Although stable packaging cell lines facilitate production the of RVVs [7], the transcriptional activity and replication of endogenous murine retrotransposons [8][9][10]49] raise biosafety concerns regarding co-packaging of endogenous murine retrotransposons along with clinicalgrade g-RVVs. Indeed, earlier pre-clinical studies reported on efficient co-packaging of the murine VL30 retrotransposon with g-RVVs generated in various murine stable packaging cells. [11][12][13][14][15]17] VL30 is a nonautonomous LTR-retrotransposon. Markopoulos et al. [50] identified 372 VL30 sequences in the mouse genome. The murine VL30 sequences include 86 full-length genomes comprising 2 LTRs and all the cis elements required for retroviral replication, including a primerbinding site (PBS), a packaging signal and a polypurine tract (PPT). VL30 promoter/enhancer sequences in the 5' LTR contain multiple transcription factor binding sites, which regulate VL30 mRNA expression and potentially exert transcriptional cis effects on neighboring host genes. [49][50][51] VL30 genomes do not encodes functional proteins. [19] Thus, the entire VL30 life cycle is dependent on either exogenous or endogenous retroviral proteins. [19,[50][51][52] However, VL30 mRNAs function as lncRNAs, most of which contain two RNA motifs, which efficiently bind and alter interaction of the human polypyrimidine tract-binding protein-associated splicing factor (PSF) with its natural DNA and RNA target sites. [20][21][22] PSF belongs to the Drosophila behavior/human splicing (DBHS) protein family and interacts with nuclear and cytoplasmic proteins, as well as with RNA and DNA target sequences. Nucleic acids/PSF interactions are mediated by a DNA binding domain (DBD) and two RNA binding domains in the PSF protein. [26,53] PSF is a multifunctional protein involved in major physiological and pathological pathways, including oncogenesis, DNA repair, RNA processing, cytokine release, viral infection, and neurodegeneration [20,26,[53][54][55][56] In mammalian cells, PSF is one of three proteins whose association with the scaffold lncRNA NEAT1_2 initiates a liquid-liquid phase separation process and the formation of membrane-less nuclear paraspeckles organelles, which regulate gene expression via several mechanisms, including sequestration of nucleoplasmic proteins and RNA molecules. An increase in paraspeckle number or size secondary to overexpression of the NEAT1_2 mRNA further recruits paraspeckles proteins, diminishes their nucleoplasmic concentration, and alters their ability to regulate the expression of genes involved in major physiological pathways. [57,58] VL30 and NEAT1_2 are not the only RNA molecules with which PSF interacts. A study by Song et al. demonstrated PSF-binding to four human mRNAs, which are overexpressed in cancer cells. [20] The ability of PSF/VL30 mRNA interaction to promote metastasis of human melanoma cells in immunodeficient mice indicated on the oncogenic potential of VL30 and other PSF-binding ncRNAs. Garen et al. outlined a molecular model of ncRNA/tumor suppressor protein (TSP) complex-induced tumorigenesis. [59,60] In this model, PSF-like TSPs bind to and inhibit transcription of protooncogenes. Binding of PSF to lncRNAs, such as VL30, results in PSF dissociation from its genomic target sequences and consequent activation of transcriptionally suppressed protooncogenes. However, various experimental systems demonstrated that the direction and mechanisms by which PSF/ncRNA complexes affect oncogenic pathways are cell type-dependent. [61][62][63][64][65] Similarly, notwithstanding the role of NEAT1 and PSF in inflammation and activation of the innate immune response [24,[66][67][68][69], paraspeckles' effects on viral infection are pathogen-specific. [24,57,64,[70][71][72][73][74][75] Based on the multiple mechanism by which PSF mediates its functions, it is difficult to predict the effects of VL30 mRNA expression on inflammatory (e.g., cytokine release) or oncogenic pathways in human T-cells. However, the oncogenic potential of PSF-binding mRNAs was considered in the wake of an earlier gene therapy pre-clinical study demonstrating the presence of VL30 genomes in lymphoma cells following bone marrow transplantation of g-RVV-transduced simian hematopoietic stem cells. [16] In this study, the abovementioned VL30 genome containing lymphoma cells did not express VL30 mRNAs, which suggested that additional mechanisms of VL30-mediated insertional mutagenesis can contribute to the oncogenic potential of VL30 genomes. Reverse-transcribed VL30 genomes preferentially integrate in proximity to transcriptional start sites, and their distribution among mouse chromosomes is not random. [50] VL30-mediated insertional mutagenesis can alter host gene expression via various mechanisms, including a) directly disrupting of host regulatory and protein-encoding sequences, b) transcriptionally activating of neighboring host genes via transcription factor-binding sites in the VL30 LTRs [50,51,[76][77][78], and c) spreading hostmediated epigenetic silencing from the VL30 LTR to neighboring genes. [79][80][81] The integration of transcriptionally active VL30 genomes into patients' chromatin raises an additional biosafety concern regarding the possibility of emerging novel retroviruses following recombination between VL30 genomes and either endogenous or exogenous retroviruses. Genomic analysis of the transforming and replication-defective Kirsten and Harvey murine sarcoma viruses (Ki-MSV and Ha-MSV, respectively) hints at the recombinational potential of VL30. Evolved by a series of recombination events. The genomes of these viruses comprise sequences from three sources: the rat VL30, the Moloney murine leukemia virus LTRs, and the Kirsten and Harvey viral ras genes, respectively. [33,34,82] In a different study, Itin et al. identified DNA recombinants comprising the VL30 LTRs and sequences with homology to the Murine Leukemia Virus (MuLV) gag and pol genes. [28] Several VL30-specific factors potentially contribute to the risk of emerging novel VL30-based recombinants, these factors include a) the presence of multiple short sequences in the human genome that exhibit high identity to sequences in VL30 genomes and b) packaging of VL30 genomes into productive human retroviral particles. Under natural conditions, sequence and structural differences between cis regulatory elements involved in all the steps of the human retroviral (e.g., HIV-1and HTLV) and VL30 life cycle minimize the risk of horizontal transfer of murine VL30 genomes by HIV-1 particles. For instance, there is no homology between the PBS sequences of VL30 strains (which are mostly complementary to the t-RNA Gly ) and the sequence of the HIV-1 PBS (which is complementary to the t-RNA Lys ). Similarly, the 3' polypurine tract (PPT) of the VL30 shows no sequence homology with that of the HIV-1. [18,19,49,50,83] Furthermore, there is no sequence homology between the VL30 and the HIV-1 packaging signals. Notwithstanding the lack of sequence homology between key VL30 and the HIV-1 cis elements, in this study, for the first time, we showed reverse transcription-dependent delivery of transcriptionally active VL30 genomes to naïve human cells by HIV-1 vector particles. However, the loss of HIV-1 vector-delivered VL30 genomes in replicating cells (following 4 passages in culture) suggested that reverse-transcribed VL30 DNA failed to integrate into transduced cells' chromatin. This phenomenon can be attributed to incompatibility between the VL30 att sites and the HIV-1 integrase [29] or/and to incompatibility between the VL30 polypurine tract (PPT) and the HIV-1 reverse transcriptase, which may alter the last step of reverse transcription and consequently leads to the formation of mostly episomal single-LTR circles (which, unlike fully reverse-transcribed linear double-stranded genomes, cannot serve as an integration template). [84] This notion is supported by an earlier study demonstrating the delivery of productive gretroviral vectors by HIV-1 particles which, similar to the abovementioned VL30 genomes, failed to integrate. In this study, genomic analysis of g-retroviral vector genomes (following delivery by HIV-1 vector particles) demonstrated the presence of episomal single-LTR circles and no linear vector forms. [30] Importantly, incorporation of HIV sequences comprising rev-response element (RRE) to g-retroviral vector genomes significantly increased their titers following packaging by HIV-1 particles. Premised on these findings, as well as the natural reversion rate of mutated HIV-1 genomes [85,86] and the VL30 recombinational potential [28,33,34], it is possible to speculate that in the presence of active HIV-1 genomes, novel VL30 recombinants can evolve to maximize their transfer by HIV-1 particles. This theoretical scenario raises biosafety concerns regarding mobilization of VL30 genomes within CAR-T treated patients and in the worst scenario within the treated patients' community. Additional consideration should be given to the fact that PFSmediated alternative splicing is central to T-cell activation. Thus, there is a possibility that CAR-T cell activation (via the CAR CD28-costimulatory domain) may be altered following binding/sequestration of nuclear PFS by VL30 mRNA. [87][88][89] Notwithstanding these biosafety concerns, more than a thousand patients with incurable oncologic diseases have been successfully treated with CAR-T cells. [90,91] To date, not a single clinical report has described proliferative abnormalities that could be attributed to the inadvertent contamination of therapeutic T-cells with VL30 genomes. [38,40,91] Theoretically, it is remotely possible that in contrast to the PG13 producer cells characterized in this study, specific vector producer cell clones that were individually isolated for specific clinical applications did not secrete VL30 genomes-comprising g-retroviral particles. Furthermore, this study did not characterize the presence of VL30 genomes in patient-administered CAR-T cells. Thus, the level of VL30 mRNA expression in patient-transplanted T-cells and the half-life of VL30 genomecontaining T-cells has not been evaluated. These unknown variables and the clinical history of a large number of CAR-T cells-treated patients suggest that there are no imminent biosafety risks associated with potentially VL30 genome-containing CAR-T cells. Importantly, short-and longterm biosafety concerns associated with the production of therapeutic g-retroviral vector carrying CAR-T cells in murine packaging cell lines could be avoided by using human packaging cell lines to generate the same CAR-carrying g-retroviral vectors. Importantly, this study underscores the importance of addressing the potential biosafety risks associated with the transfer of non-human endogenous retroviruses and potentially nonautonomous retrotransposons to patients undergoing novel therapeutic procedures. [92][93][94][95][96]
Isolation of single-cell 293T clones containing VL30 genomes 293T cells were transduced with the lentiviral vector pTK1261, from which the firefly luciferase cDNA and the fusion GFP/blasticidin marker gene were expressed under the control of a CMV promoter and the encephalomyelitis virus internal ribosome entry site (IRES), respectively. Transduced cells were selected for blasticidin resistance in the presence of 50µg/ml blasticidin (293T-1261). Blasticidin resistant 293T cells were co-cultured with PG13 cells. At confluency, the co-culture (293T-1261/PG13) cell population was passaged, and PG13 cells were eliminated in the presence of 50µg/ml blasticidin. The abovementioned process of co-culturing was repeated seven times, after which single cell clones of blasticidin-resistant 293T cells were isolated. The absences of contaminating PG13 cells was confirmed by the lack of PCR amplification of the mouse GAPDH gene. GCN in the abovementioned isolated single-cell clones was determined by qPCR as described below.
Production CAR carrying g-RVV from a newly established stable producer cell line A stock of Eco-pseudotyped SFG.iC9.GD2.CAR.IL15 g-RVV was produced by transient transfection of the ecotropic ΦNX-Eco packaging cell line (American Type Culture Collection product CRL-3214; ATCC, Manassas, VA) with the vector-expression cassette. A heterogenous stable producer cell line was generated by repeated transduction of the murine gibbon ape leukemia virus (GalV) envelope-expressing PG13 packaging cell line (# CRL-10686™, ATCC, Manassas, VA) with the abovementioned Eco-pseudotyped SFG.iC9.GD2 vector. To enrich for stable producer cells with high VCN, the abovementioned heterogenous population of vector producing cells was immuno-stained with an anti-idiotype antibody specific for the GD2 CAR [31] high transgene expressing cells were sorted using the BD Jazz cell sorter. Following limiting dilution single-cell clones of vector-producing cells were isolated and functionally screened for the highest biological titer by qPCR. The highest vector producing clone was expanded for generation of the Master Cell Bank (MCB) for production of GalV envelope-pseudotyped retroviral particles carrying the iC9.GD2.CAR.IL15 expression cassette. For each lot of g-RVV supernatant, cells from the abovementioned MCB were thawed and expanded in the iCellis Nano bench-top bioreactor, (Pall, Inc). The iCellis Nano is a fixed bed bioreactor using fiber material for the attachment of the attachment of adherent vector producer cells. The system was employed to provide continuous monitoring of dissolved oxygen, pH, and temperature. The system provided continuous circulation of culture media along with proper gas mixture of O2 and CO2. pH was additionally controlled by the addition of base as required. After seeding onto the fiber bed of the iCellis Nano, the cells are cultured for 2 -3 days to reach near confluency. At this point, the media was replaced with fresh media, and following overnight culture, the supernatant was harvested into a transfer bag, and fresh media was added to the system. This supernatant is termed "Day 1" harvest. This process was repeated four additional times resulting in a total of five days of supernatant harvest. After harvest of the fifth supernatant, the producer cells were removed from the bioreactor using enzymatic digestion (TrypLE) and washing with PBS. The harvested cells were counted and samples taken for QC studies. The MCB, end of production cells and viral supernatant were tested following FDA recommendation for sterility and absence of replication competent retroviral particles.

Lentiviral and g-retroviral vectors
The construction of the non-SIN g-RVV, SFG.iC9.GD2.CAR.IL15 was described earlier. [31] pTK1261 was constructed by cloning a BglII/BamHI DNA fragment containing the firefly luciferase cDNA into a BamHI site in the lentiviral vector pTK642. [97] pTK2229 was constructed by cloning an AfeI/HpaI DNA fragment comprising a VL30 sequence from nucleotides 2140 to 2769 of GeneBank: AF486451.1 into a PshAI site in a lentiviral vector comprising a CMV promoterregulated expression cassette (encoding the mCherry-T2A-Puromycin selection marker) in opposite orientation to the LTRs. The pTK1808 and pTK2151 vectors were purchased from Addgene (Addgene, Watertown, Cat# 14088 and 10668, respectively). Both vectors are premised on the g-retroviral vector pBabe-puro and carry either the SV40 large-T antigen or the green fluorescence protein (GFP) cDNA under the control of the vector 5' LTR, respectively. Note that downstream to the SV40 large-T antigen cDNA, the pTK1808 vector also contains the puromycin resistance cDNA under the control of the SV40 promoter.

Production of viral vectors by transient transfection
All vector particles were VSV-g pseudotyped and produced in 293T cells using the transient threeplasmid calcium phosphate transfection method as described earlier. [98,99] In brief, the secondgeneration lentiviral vector packaging cassette ΔNRF [100] or the MLV Gag/Pol expression cassette (a kind gift from Dr. Nikunj Somia at the University of Minnesota) were transiently transfected into 293T cells along with the VSV-G envelope expression cassette and the relevant vector construct. Vector particles in conditioned media were harvested ~60 hours after transfection and filtered through a 0.45-µm syringe filter.
Analysis of VL30 genome-and g-retroviral vector copy number (GCN and VCN, respectively) Genomic DNA samples were extracted by the DNeasy Blood & Tissue Kit (69506, QIAGEN, Hilden, Germany). To eliminate potential contamination with carried-over transfected plasmid DNA, all genomic DNA samples (except DNA samples extracted from cultured human T-cells) were digested with the DpnI restriction enzyme. GCN and VCN were determined by quantitative real-time polymerase chain reaction (qPCR) using the QuantStudio TM 3 system (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA). To measure VL30 GCN, a reference cell population containing a single-copy of a VL30 sequence was established (293T-2229 cells). Specifically, 293T cells were transduced with the lentiviral vector pTK2229 (at MOI <0.001) and selected for puromycin resistance in the presence of 5 µg/ml of puromycin (P8833, MilliporeSigma, St. Louis, MO). DNA samples extracted from the abovementioned heterogenous population of puromycin resistant 293T cells served to establish a reference DNA standard-curve to measure VL30 VCN by qPCR using a primer/probe set (Integrated DNA Technologies, Inc., Coralville, Iowa) comprising a forward primer 5'-CCTTGACCAGAAGCCACTATG-3', a reverse primer 5'-TCAGAGATTGGGACCCTGAA-3', and a 6-FAM TM -conjugated probe 5'-TGTAAGATGGCCTGCTTGT CTGCA-3'. To measure the VCN of g-RVVs (expressing either a CAR or the GFP cDNA), a reference cell population comprising a single-copy of a g-RVV genome was established (293T-1808 cells).
Specifically, 293T cells were transduced with the g-RVV l vector pTK1808 (at MOI <0.001) and selected for puromycin resistance in the presence of 5 µg/ml of puromycin (MilliporeSigma, St. Louis, MO). DNA samples extracted from the abovementioned heterogenous population of puromycin-resistant 293T cells served to establish a reference DNA standard-curve to measure g-RVV VCN by qPCR using a primer/probe set (Integrated DNA Technologies, Inc., Coralville, Iowa) comprising a forward primer 5'-CGCTGACGGGTAGTCAATC-3', a reverse primer 5'-GGGTACCCGTGTATCCAATAAA-3' and a 6-FAM TM probe 5'-ACTTGTGGTCTCGCTGTT CCTTGG-3'. qPCR analysis of the endogenous human RNaseP, which served as an internal reference control, was premised on a commercial human RNaseP primer/probe set (443328, Applied Biosystems, Thermo Fisher Scientific, Waltham, MA). PCR amplification of the mouse GAPDH gene was premised on a primer/probe set from Hoffmann-La Roche Ltd, Basel, Switzerland (Universal ProbeLibrary Mouse GAPD Gene Assay, 05046211001). qPCR was performed with the ABsolute qPCR ROX Mix (AB-1138/B, Applied Biosystems, Thermo Fisher Scientific, Waltham, MA) under the following conditions: 95°C for 15 min, and then 40 cycles of 95°C for 15 sec and 60°C for 1 min.

RNA isolation and qRT-PCR
RNA was isolated using an RNeasy® Plus Mini Kit (QIAGEN, Hilden, Germany) and converted to cDNA using a QuantiTectâ Reverse Transcription Kit (QIAGEN, Hilden, Germany). qRT-PCR of VL30 and g-RVV mRNA was premised on the same primer/probe sets that were described above. The qRT-PCR assay for the human ACTB mRNA, which served as an internal reference control, was premised on a commercial set of primers/probe (Hs.PT.39a.22214847) conjugated with HEX TM at the 5' end (Integrated DNA Technologies, Inc., Coralville, Iowa). All PCR results were analyzed with Prism 9 software (GraphPad Software, San Diego, CA).

Statistical analysis
VL30 genome copy number (GCN) and g-RVV vector copy number (VCN) in human T-cells transduced with CAR-expressing g-RVV were calculated as the average of 3 technical replicates. The association between VL30 GCN and g-RVV in primary human CAR-T cells was estimated by Pearson correlation and tested with linear regression analysis. Statistical analyses employed to characterize the significance of various treatments' effects on VL30 GCN and g-RVV VCN are outlined in the figure legends.

Disclosure
TK is an inventor of PPT-deleted lentiviral vectors and of integration defective lentiviral vector production technologies, which are owned by the University of North Carolina. Some of these technologies are licensed to a commercial entity. GD is a paid consultant for Bellicum Pharmaceuticals, Tessa Therapeutics and Catamaran. BS is supported by Bluebirdbio, Bellicum Pharmaceutical, Cell Medica, Tessa therapeutics and is a paid consultant for Tessa Therapeutics. The other authors declare that they have no conflict of interest.

A.
B.
- The experiment was performed in triplicate. C. Bar graph showing VCN of the g-RVV pTK2151 in 293T cells following exposure to conditioned media collected from either PG13 cells transduced with the g-RVV pTK215, or 293T cells transiently transfected with the pTK2151 vector cassette, a VSV-g envelope and an g-RVV packaging cassettes. Exposure to the abovementioned conditioned media was done either in the presence or absence of the reverse-transcriptase inhibitor AZT (10µM). The experiment was performed in triplicate Significance of the AZT effect on VL30 GCN and pTK2151 VCN was determined by 2-way ANOVA, *P≤0.05, **P≤0.01 Table 1. The effects of a reverse-transcriptase inhibitor of VL30 transduction. A. The table presents the raw data described in Figure 2-A. The conditioned media producing cell lines are outlined. The absence or presence of AZT (10µM) at the time of exposure to the abovementioned conditioned media is indicated by -and + signs, respectively. The copy number of VL30 genomes (GCN) in the abovementioned 293Ts is indicated. ND indicates VCN levels that were lower than the lower detection level by qPCR. The values of standard deviation (std) are shown. B. The table presents the raw data described in Figure 2-B. The conditioned media producing cell lines are outlined. The absence or presence of AZT (10µM) at the time of exposure to the abovementioned conditioned media is indicated by -and + signs, respectively. VL30 GCN in the abovementioned 293Ts is indicated. ND indicates VCN levels that were lower than the detection level obtained through qPCR. The values of standard deviation (std) are shown. Table 2. Copy number of VL30 genomes in single-cell clones of HEK 293T cells following co-culturing with PG13 cells     To test the hypothesis that human pathogens, including HIV-1, can potentially transfer VL30 genomes, 293T cell clones 6 and 7 (VL30 genome copy number of 5.17 and 2.68, respectively) were transiently transfected with the HIV-1 vector packaging and the VSV-G envelope-expression cassettes. Vector particles were employed on naïve 293T cells either in the presence or absence of 10µM AZT. DNA was extracted from treated 293T cells and VL30 genome copy numb was determined by qPCR. A. Graph bar showing VL30 genome copy number in naive 293T cells exposed to lentiviral vector particles generated in cell clones 6 and 7, either in the presence or absence of 10 µM AZT. DNA samples extracted from 293T cells comprising the lentiviral vectors pTK1261 or pTK2229 served as negative and positive controls, respectively. Note that in the absence of AZT, the level of VL30 genome copy number in naïve 293T cells that were exposed to lentiviral vector particles generated in clone 6 and 7 (0.0429 ±0.0094 and 0.0115±0.0022, respectively) correlates with VL30 genome copy numbers in the respective vector producing single cell clones ( Figure 3). The significant reduction in VL30 genome copy number in AZT treated 293T cells indicates that the observed transfer of VL30 genomes was reverse-transcription dependent. PCR amplification products of the endogenous RNaseP gene served as loading controls. P-values were determined by the 2-way ANOVA test. The experiment was performed in triplicate. B. Electrophoresis analysis of PCR products following amplification of DNA samples extracted from 293T cells transduced by HIV-1 particles generated in cell clones 6 and 7. DNA samples extracted from 293T cells transduced with the lentiviral vectors pTK1261 and pTK2229 served as negative and positive controls, respectively. Table 3. Data of the qPCR assays described in Figure 4 Table 5. Genome copy number (GCN) and vector copy number (VCN) of VL30 and g-RVV (respectively), in DNA samples extracted from primary human Tcells as determined by qPCR analysis. DNA samples 1, 2, 5, 6, 7, 11, 12 and 13 were extracted from human T-cells treated with the g-RVV SFG.iC9.GD2.CAR.IL15. DNA of samples 3,4,8,9 and 10 were extracted from naïve primary human T-cells and served as biological negative controls. DNA extracted from naïve 293T cells served as technical negative control (-cont). ND indicates that the level of targeted sequences (VL30 and g-RVV) in the tested DNA sample were under the level of detection.
g-RVV copy number per cell Figure 6. qPCR-based analysis of VL30 genome copy number (GCN), and g-RVV vector copy number (VCN) in primary human T-cells. Naïve primary human T-cells were treated with the chimeric antigen receptor (CAR)-carrying g-RVV SFG.iC9.GD2.CAR.IL15 (samples 1, 2, 5, 6, 7, 11, 12 and 13). The vector was produced in the stable packaging cell line PG13. Samples 3,4,8,9 and 10 were not exposed to the abovementioned CAR-carrying g-RVV and served as biologic negative controls. DNA samples extracted from the abovementioned control and vector-treated cells were used to determine GCN and VCN of the murine endogenous retrotransposon VL30 and the g-RVV vector, respectively. DNA from naïve 293T cells served as a technical negative control. DNA from 293T cells transduced with the lentiviral vectors pTK2229 and pTK2151 served as positive control for VL30 and g-RVV genome amplification, respectively. Amplification of the endogenous human gene hRNaseP served as loading control. The assay was performed in technical triplicate. A. A Physical map depicting the structure of the CAR-carrying g-RVV, SFG.iC9.GD2.CAR.IL15. The vector's non-self-inactivating (non-SIN) long terminal repeats (LTRs) and the packaging signal y are shown. The inducible caspase-9 suicide gene, Ic9 is shown. The self-cleavable peptide from the thosea asigna virus (T2A) and the equine rhinitis virus (E2A) are shown. Sequences encoding the CAR including the single-chain variable fragment (scFv1) directed to the NB-antigen GD2 of the disialoganglioside GD2 (14g2a), the CD8a stalk and transmembrane domain, the CD28 intracellular domain and the CD3z chain are shown. The human IL15 cDNA containing sequence is shown. B. Graph bar showing GCN of VL30 in the abovementioned primary human T-cells. The experiment was performed in technical triplicate. C. Graph bar showing VCN of the g-RVV SFG.iC9.GD2.CAR.IL15 in the abovementioned primary human T-cells. The experiment was performed in technical triplicate. D. Gel electrophoresis analysis of DNA amplification products of VL30, g-RVV and hRNaseP sequences generated in the course of the above gPCR assay. DNA of naïve 293T cells (N) served as technical negative control for amplification of VL30 and g-RVV sequences. DNA from 293T cells transduced with the lentiviral vectors pTK2229 and pTK2151 served as positive controls (P) for VL30 and g-RVV genome amplification, respectively. E. Significant correlation between VL30 and g-RVV genome copy number in human T-cells transduced with CAR-expressing g-RVV. Linear regression graph demonstrating the relation between VL30 and g-RVV genome copy number. The R value is indicated.

2^-ΔCt of VL30
A. B. Table 6.     Table 7. continued Table 7. continued Table 7. continued Table 7. Sequence homologies between various VL30 genomes and the human genome. The Genome Browser of the University of California Santa Cruz Genomic Institute (https://genome.ucsc.edu/) was employed to identify sequence homologies between various VL30 genomes and the human genome. The accession numbers comprising the sequences of the various VL30 genomes, which were employed in the sequence VL30/human homology screen are indicated. The position of the first and last (start and end respectively) VL30 nucleotides in the homologous sequence are shown. VL30 nucleotides that matched their respective nucleotides in homologous human sequences are shown in capital letters. The number of human chromosomes containing sequences with homology to VL30 genome sequences are indicated. The orientations of the relevant homologous strands are shown. The position of the first and last (start and end respectively) nucleotides in the human chromosomes with homology to the relevant VL30 sequence are shown. The repetitive sequences in the human chromosomes with homology to the abovementioned VL30 genomes are defined.